Coplanar electrowetting-induced stirring as a tool to manipulate biological samples in lubricated digital microfluidics. Impact of ambient phase on drop internal flow pattern Laurent Davoust, Yves Fouillet, Rachid Malk, and Johannes Theisen Citation: Biomicrofluidics 7, 044104 (2013); doi: 10.1063/1.4817006 View online: http://dx.doi.org/10.1063/1.4817006 View Table of Contents: http://bmf.aip.org/resource/1/BIOMGB/v7/i4 Published by the AIP Publishing LLC.

Additional information on Biomicrofluidics Journal Homepage: http://bmf.aip.org/ Journal Information: http://bmf.aip.org/about/about_the_journal Top downloads: http://bmf.aip.org/features/most_downloaded Information for Authors: http://bmf.aip.org/authors

Downloaded 30 Aug 2013 to 152.14.136.96. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://bmf.aip.org/about/rights_and_permissions

BIOMICROFLUIDICS 7, 044104 (2013)

Coplanar electrowetting-induced stirring as a tool to manipulate biological samples in lubricated digital microfluidics. Impact of ambient phase on drop internal flow patterna) Laurent Davoust,1,b) Yves Fouillet,2 Rachid Malk,2 and Johannes Theisen3 1

Grenoble Institute of Technology (Grenoble-INP), Materials and Processes Science and Engineering Laboratory (SIMAP), BP 75, 38402 Saint Martin d’He`res, France 2 CEA-LETI-MINATEC, 38041 Grenoble cedex 9, France 3 University of Grenoble (UJF), Laboratory of Geophysical and Industrial Fluid Flows (LEGI), BP 53, 38041 Grenoble, France (Received 28 April 2013; accepted 17 July 2013; published online 25 July 2013)

Oscillating electrowetting on dielectrics (EWOD) with coplanar electrodes is investigated in this paper as a way to provide efficient stirring within a drop with biological content. A supporting model inspired from Ko et al. [Appl. Phys. Lett. 94, 194102 (2009)] is proposed allowing to interpret oscillating EWOD-induced drop internal flow as the result of a current streaming along the drop surface deformed by capillary waves. Current streaming behaves essentially as a surface flow generator and the momentum it sustains within the (viscous) drop is even more significant as the surface to volume ratio is small. With the circular electrode pair considered in this paper, oscillating EWOD sustains toroidal vortical flows when the experiments are conducted with aqueous drops in air as ambient phase. But when oil is used as ambient phase, it is demonstrated that the presence of an electrode gap is responsible for a change in drop shape: a pinch-off at the electrode gap yields a peanut-shaped drop and a symmetry break-up of the EWOD-induced flow pattern. Viscosity of oil is also responsible for promoting an efficient damping of the capillary waves which populate the surface of the actuated drop. As a result, the capillary network switches from one standing wave to two superimposed traveling waves of different mechanical energy, provided that actuation frequency is large enough, for instance, as large as the one commonly used in electrowetting applications (f  500 Hz and beyond). Special emphasis is put on stirring of biological samples. As a typical application, it is demonstrated how beads or cell clusters can be focused under flow either at mid-height of the drop or near the wetting plane, depending on how the nature of the capillary waves is (standing or traveling), and therefore, depending on the actuation frequency (150 Hz–1 KHz). C 2013 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4817006] V I. INTRODUCTION

Overcoming diffusion of molecules and mixing of reagents in biotechnological applications and assays is challenging, the aim being to ensure rapid and efficient assays.2 New ways of realizing stirring in miniaturized systems like lab-on-chips are worthy to be explored. A second application of large importance at the scale of micro-systems is the preparation of biological samples with the purpose to extract biologically functionalized beads or clusters of cells, for instance. As a technological breakthrough, ElectroWetting On Dielectric (EWOD) stands as a new driving mechanism in micro-systems like variable focal lenses,3 displays4 or digital lab-on-a-chip devices for a)

Paper submitted as part of the 3rd European Conference on Microfluidics (Guest Editors: J. Brandner, S. Colin, G. L. Morini). The Conference was held in Heidelberg, Germany, December 3–5, 2012. b) Author to whom correspondence should be addressed. Electronic mail: [email protected]. 1932-1058/2013/7(4)/044104/14/$30.00

7, 044104-1

C 2013 AIP Publishing LLC V

Downloaded 30 Aug 2013 to 152.14.136.96. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://bmf.aip.org/about/rights_and_permissions

044104-2

Davoust et al.

Biomicrofluidics 7, 044104 (2013)

clinical diagnostics5 or biological environmental monitoring.6 Used in AC mode at a frequency, f  1 KHz or more, EWOD is used as a common means to move, merge or break-up drops.7 EWOD can also be used at a lower frequency in a regime referred to as oscillating EWOD (f  100 Hz). This permits to induce shape oscillations of a sessile drop which in return generates a Stokes drift along its liquid surface. Subsequently, oscillating EWOD has been proposed as a mixing promoter in two popular configurations of electrowetting: the parallel-plate7,8 and needle9–12 configurations. In this paper, oscillating EWOD-induced drop stirring is provided with a coplanar electrode open configuration which reveals itself to be easily integrable into state-of-the-art digital labon-chips. As a consequence, EWOD-induced mixing enhancement is experimentally promoted without the presence of a plunging needle as a counter-electrode. In presence of a needle, the electrostatic map is changed compared to the one of the coplanar configuration,13 at least for a large enough frequency (f  10–100 KHz), and capillary effects due to moving meniscus along the needle should be taken into account. Coplanar geometry allows us to compare in best conditions our experimental results with current streaming theory available in the literature either at bubble scale,1 or at the scale of a droplet in air as ambient phase10 or, eventually at the scale of a droplet in oil as ambient phase (see Ko et al.14 for the moment but coupling with electrostatics is not considered). Typically, when the frequency is small enough and the ambient phase is air, coplanar EWOD results in a single spherical standing wave along the drop surface with capillary resonance15 properly achieved at the apex.16,17 In a first part of this paper, it is demonstrated how EWOD-induced stirring can be considered as a liquid surface flow generator; the surface momentum it generates is dissipated within the viscous drop; the smaller the drop is, the more efficient this surface flow generator is. A second part is devoted to a symmetry break-up of the internal flow pattern, jointly induced by the presence of an electrode gap and the use of oil as alternative ambient phase. A last part is devoted to biological applications of oscillating EWOD used to focus colloidal beads and biological cells (U373B) at drop scale. II. MATERIALS AND METHODS

Technological steps are performed on a 200 mm silicon wafer and EWOD chips are composed of three main layers: the electrodes (200 nm thick AlCu); the dielectric layer (600 nm thick Si3N4); and, finally, the hydrophobic layer (1 lm thick SiOC). These materials enable to integrate the whole fabrication process into cleanrooms with no compromise on EWOD performance and a good reproducibility during assays. More details of the technology are given in Malk et al.18 The chip design used here consists of two half disks shaped coplanar electrodes (radius ¼ 2 mm) separated by a 3 lm-gap (as regards experiments with air as ambient phase) or a 100 lm-gap (as regards experiments with oil as ambient phase, Figs. 1 and 2). A first advantage of this coplanar electrode configuration—compared with the classic needle electrode configuration—is that droplet deformation is not disturbed by meniscus effect due to the presence of a needle electrode. Consequently, top views with image analysis are made possible. A second advantage lies on the circular geometry of the two facing electrodes which enhances isotropic electrowetting of the sessile drop and makes analytical modeling possible. AC electric signal is generated from a generator (Yokogawa FG120) and a home-made amplifier. A CCD camera (Pixelfly 200 XS) and a zoom lens are used for image acquisition. The camera is placed so as to visualize droplet profiles (Fig. 2). A LED is used as a stroboscopic lighting source placed behind the droplet in order to create a backlight and to minimize light reflection. The LED is plugged to the same generator. The frequencies of AC actuation and stroboscopic lighting are equal, and by shifting the phase lag between both signals, the instantaneous shape of the droplet can be clearly imaged. Time-dependent droplet profiles are thus acquired and analyzed using home-made software. Drop convective flow is simultaneously analyzed by imaging fluorescent beads (mean diameter: 10 lm and mean density: 1.05) and using a software dedicated to micro-PIV (Particle

Downloaded 30 Aug 2013 to 152.14.136.96. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://bmf.aip.org/about/rights_and_permissions

044104-3

Davoust et al.

Biomicrofluidics 7, 044104 (2013)

FIG. 1. Design of the electrode pair under consideration.

Image Velocimetry). The laser sheet used to induce bead fluorescence is provided by a laser source (Melles Griot: 5mW, k ¼ 470 nm) and a series of two cylindrical lenses (thickness of laser sheet at drop position: 100 lm). A 532 nm filter is fitted to the zoom lens to detect fluorescence. It is thus possible to characterize simultaneously both the flow and the droplet oscillations at different frequencies of the applied voltage. With the beads and the lighting source used in the experiments, the flow that is observed is the drop surface flow. For the observation of internal flows, use is made of the laser sheet. Most of the experiments are realized with a 1.5 ll drop of Phosphate Buffered Saline (PBS) deposited on the lubricated chip surface with a micropipette. Depending on the experiments, air or silicone oil (RT5, Paragon Scientific, dynamical viscosity: l ¼ 4,9 mPa s) can be

FIG. 2. Experimental setup and array of EWOD chips in the upper frame.

Downloaded 30 Aug 2013 to 152.14.136.96. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://bmf.aip.org/about/rights_and_permissions

044104-4

Davoust et al.

Biomicrofluidics 7, 044104 (2013)

TABLE I. Wetting data in static conditions with the EWOD chip considered in this study. PBS drop in:

Air

Oil

None EWOD None EWOD

Surface tension (N/m) Contact angle

72 107

28 158

None EWOD

Contact angle hysteresis

5  7

1 KHz): Ro ¼ ð3V 2p Þ , with Vl, the liquid volume deposited. For a standing wavy surface deformed with a kth-eigenmode, the time-dependent velocity potential in the inner of the drop writes as:23,24

Downloaded 30 Aug 2013 to 152.14.136.96. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://bmf.aip.org/about/rights_and_permissions

044104-6

Davoust et al.

Biomicrofluidics 7, 044104 (2013)

Ui ¼ 

Ak nk Ro r k Pk ðcosðhÞÞcosðnk t þ wk Þ: k Rko

Hence, it is possible to calculate the normal component of the surface velocity, 1 @Ui  u1 k ¼  Ro @h r¼Ro , as follows: 1 u1 k ¼ Uk cosðnk t þ wk Þ;

where Uk1 ¼

Ak nk Pk ðcosðhÞÞ; k

is nothing but the normal velocity of the standing wave. Considering a Strouhal number large enough, S¼

nk Ro  1; Uk1

and therefore, the theory of steady streaming,25,26 it is demonstrated, in a way similar to Ko et al. for electrowetting-driven bubble oscillations,1 that a steady component of the velocity, Uk s , referred to as streaming current, is provided at the boundary of (and beyond) a Stokes layer staying along the inner of the moving drop surface and writes as: Uks ¼ 

3 d ½ðUk1 Þ2 ; 8R0 nk dh

or in terms of Legendre polynomials: Uks

3A2 n d ¼  k 2k 8R0 k dh

"

dPk ðcos hÞ dh

2 # :

(1)

In our case, the thickness of the Stokes layer is much thinner than the drop radius, qffiffiffiffiffi ds ¼ Oð qng k Þ  Ro , with g and q, the viscosity and the density of the drop, respectively (Table III). And consequently, the streaming velocity can be considered as a surface boundary condition for the drop convective flow, which therefore may be calculated inside a permanent half spherical drop (time-averaged shape). This is clearly the reason why EWOD-induced streaming current can be thought of as an efficient flow generator at micro-scale.

C. Numerical calculations and comparison with experiments on PBS drops in air q ðU 1 Þ2

From the literature,27 the streaming Reynolds number, Rsk ¼ nkkg , depends on the oscillating mode, k ¼ 2,4,.etc. In our case, its value is never negligibly small (Rsk  10  100, see, e.g., Table III) and consequently, the drop flow cannot be considered as creeping. Hence, numerical

TABLE III. Typical values of the parameters associated to streaming theory for water drops in air (drop volume: 1.5 ll). Mechanical nk frequency: 2p

Ak =Ro

100 Hz

200 Hz

0.3

4  105

3.33

90

1000 Hz

2000 Hz

0.05

105

20

25

Voltage frequency: x2pk

ds ¼

qffiffiffiffiffi g qnk

S ¼ nUk R1o k

Rsk ¼

qðUk1 Þ2 gnk

Downloaded 30 Aug 2013 to 152.14.136.96. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://bmf.aip.org/about/rights_and_permissions

044104-7

Davoust et al.

Biomicrofluidics 7, 044104 (2013)

TABLE IV. Eigenmodes and resonant frequencies for PBS drops in air (drop volume: 1.5 ll). Consider also Ref. 21 for viewing resonant drop shape oscillations. Eigenmode k Resonant frequency nk (610 Hz)

2 110

4 216

6 393

8 600

10 835

12 1093

R software (based on finite element method) in order simulations are conducted under ComsolV to predict the drop flow induced by EWOD-streaming and to compare it with our velocimetry experiments performed with micro-PIV of fluorescent beads. The averaged drop shape considered for present calculations is supposed to be a half-sphere, which is not so restrictive since in most applications, AC EWOD delivers a contact angle close to p/2. All the oscillating modes are time-averaged and the expression (1) for the surface velocity is introduced as a boundary condition along the drop surface for each mode considered. In this way, drop shape oscillations are nevertheless taken into account through their impact upon the tangential momentum which is injected from the Stokes layer. Due to 2-D axisymmetry, flow calculation is performed within a meridian cross-section and two slip conditions are imposed along the vertical drop axis and the underlying substrate lubricated with a thin film of RT5 oil (thickness: 260 nm (Ref. 16)). For sake of comparison, we select experiments conducted on drops of PBS in air as ambient phase. The actuation frequency is set to values small enough to promote a standing capillary wave all along the drop surface (see Table IV for the dependence of resonant frequencies on the eigenmode k). This experimental requirement allows us to be consistent with the steady surface velocity as predicted by Eq. (1). Top and side views of the inner flow as produced under oscillating EWOD (f ¼ 110 Hz, mode k ¼ 2) are displayed in Figs. 3(a) and 3(b). Two (axisymmetric) toroidal vortices are (is) consistently made evident from imaging of fluorescent beads and for sake of comparison, streamlines are calculated within a meridian cross-section (Fig. 3(c)). In order to calculate the real locations of the tracers during micro-PIV developments, use was partly made of the image mapping method proposed by Kang et al.,28 corrected by considerations from Minor et al.29 In this way, it was possible to take into account light ray deviations due to surface curvature and jump in refractive index across the drop surface. As regards the micro-PIV calculation of tracers trajectories, different time scales must be considered: a first and small one, T1, associated to the time aperture of the camera, a second one associated to trajectory fluctuations, especially within the Stokes layer, which is of the order of the inverse of the actuation frequency, T2 ¼ n1k with T2  T1 , an intermediate one, T3, which is the time delay between two successive frames, and finally, a larger one, T4, which must be selected conveniently for averaging the series of frames. The steady velocities displayed in Fig. 3 are found after averaging a series of frames over time scale T4. The time scales T3 and T4 are set conveniently to get both statistically invariant velocities and a high enough spatial resolution. Hence, a fair agreement is demonstrated and it is worthy to mention that the velocity measured by micro-PIV at the vicinity of the drop apex is of same order as the numerical prediction ¼ 0:045 and u ’ 6:5 mm=s for whatever the relative mode amplitude is: u ’ 2 mm=s for ARk¼2 o Ak¼2 ¼ 0:06 (see Figs. 3 and 4). Ro A satisfying agreement between experiments on drops in air and numerical predictions is also found for oscillation modes k ¼ 4 and 6, for which the actuation frequency, f, is never larger than 450 Hz (Fig. 4). The qualitative agreement is remarkable as regards the streamlines (see top of Fig. 4), while it is increasingly poor when comparison is made for larger and larger modes.21 For k ¼ 6, we think that the standing nature of the capillary wave becomes questionable since the mechanical energy of the radially inwards traveling capillary wave (TCW) becomes more predominant. Beyond the critical frequency, f ¼ 450 Hz, it is no longer possible to observe toroidal flows, for instance those expected for the mode k ¼ 8: the drop flow is no more found axisymmetric and steady. And when imaging the streamlines from above, counterrotating vortices with rotation about an arbitrary axis are made evident.

Downloaded 30 Aug 2013 to 152.14.136.96. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://bmf.aip.org/about/rights_and_permissions

044104-8

Davoust et al.

Biomicrofluidics 7, 044104 (2013)

FIG. 4. Comparison (modes k ¼ 2, 4, 6) for a sessile PBS drop in air between the velocities as measured at the apex of the drop or numerically predicted from Eq. (1) with h ¼ 0. The experimental velocity is obtained from micro-PIV measurements especially performed on the small area bounded by the red rectangle displayed in the snapshots (tracers: fluorescent beads). The streamlines as computed numerically are made visible on the RHS of the snapshots. Chip with two halfcircular electrodes separated by a gap of 3 lm, RMS voltages: 60 V (k ¼ 2), 75 V (k ¼ 4) or 91.7 V (k ¼ 6). Consider also Ref. 21 for imaging of the streamlines.

D. Comments on steady internal drop flow pattern 1. Drop flows in needle electrode configuration: State of art

As already mentioned, most of the literature devoted to internal flow pattern inside a drop under EWOD focuses on the needle electrode configuration.11,30,31 In such configuration, the meniscus motion along the needle and eventually, at least at a large enough frequency (f  10–100 KHz), the subsequent distribution of the electric potential within the drop, make the flow pattern different from the one reported in our experiences on coplanar configuration. •



Air used as ambient phase In the needle electrode configuration, when air is used as ambient phase, only one toroidal vortex is reported by Ko et al.31 with a flow directed downwards (upwards) along the symmetry axis and radially outwards (inwards) along the supporting substrate if the actuation frequency is large enough, f  100 KHz (small enough, f  1 KHz). This frequency-driven flow reversal originates from electrothermal effects.10,32 Oil used as ambient phase Still for the needle electrode configuration, but when oil is used as ambient phase, a drop flow reversal is again found11,30 at a small frequency (f  1 KHz at most), compared to previous experiments in air carried out by Ko et al.31 at same frequency. External viscous dissipation enhanced by oil is suspected to damp the two radially inwards and outwards TCW which populate the drop surface. Since the radially inwards TCW is primarily excited from the actuated oscillating contact line, the radially outwards TCW is expected to vanish primarily under viscous damping. Therefore, remaining mechanical energy is associated to the radially inwards TCW which is suspected to explain the oil-induced flow reversal. It is worthy to recall here that Stokes drift induced by one single TCW is not the same as the one derived from a standing capillary wave.25 This scenario seems to be confirmed by recent numerical calculations performed by Oh et al.14

Downloaded 30 Aug 2013 to 152.14.136.96. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://bmf.aip.org/about/rights_and_permissions

044104-9

Davoust et al.

Biomicrofluidics 7, 044104 (2013)

2. Drop flows in coplanar electrodes configuration

Here, in our experiments on coplanar configuration, when air is used as ambient phase, the observed wave pattern has been found perfectly standing for eigenmodes k ¼ 2, 4, 6. Both the absence of a needle as a counter-electrode and the presence of a lubricated substrate allow the apex of the drop as well as the contact line to move freely (negligible contact angle hysteresis in lubricated conditions:

Coplanar electrowetting-induced stirring as a tool to manipulate biological samples in lubricated digital microfluidics. Impact of ambient phase on drop internal flow pattern.

Oscillating electrowetting on dielectrics (EWOD) with coplanar electrodes is investigated in this paper as a way to provide efficient stirring within ...
4MB Sizes 0 Downloads 0 Views